A market-model CAR is a measurement of the price reaction. It becomes a causal claim only with an additional design: an instrument when the regressor is a choice the firm made, or a robust covariance when the residuals are dependent. This topic states those extensions without re-teaching the core instrumental-variables algebra, then closes with a one-announcement design memo and a numerical fixture whose ratio is computed from the source formulas and is not a historical test statistic.
By the end of this note you should be able to:
- state which of Wooldridge's two instrument conditions a reduced form can test;
- assign White, a firm-level cluster sandwich, or Newey–West to a stated dependence;
- recompute the memo's three-day CAR and its large-sample standard deviation from the fixture table.
In this lesson
- Contents
- The big picture
The full lesson is for members.
Read the full lessonNew to SelfMadeQuant.com? See membership options or browse the Event Studies and Causal Measurement in Finance syllabus.